Optimal design, anti-tumour efficacy and tolerability of anti-CXCR4 antibody drug conjugates.

Antibody-drug conjugates (ADCs) are promising therapies for haematological cancers. Historically, their therapeutic benefit is due to ADC targeting of lineage-restricted antigens. The C-X-C motif chemokine receptor 4 (CXCR4) is attractive for targeted therapy of haematological cancers, given its expression in multiple tumour types and role in cancer "homing" to bone marrow. However, CXCR4 is also expressed in haematopoietic cells and other normal tissues, raising safety challenges to the development of anti-CXCR4 ADCs for cancer treatment. Here, we designed the first anti-CXCR4 ADC with favourable therapeutic index, effective in xenografts of haematopoietic cancers resistant to standard of care and anti-CXCR4 antibodies. We screened multiple ADC configurations, by varying type of linker-payload, drug-to-antibody ratio (DAR), affinity and Fc format. The optimal ADC bears a non-cleavable linker, auristatin as payload at DAR = 4 and a low affinity antibody with effector-reduced Fc. Contrary to other drugs targeting CXCR4, anti-CXCR4 ADCs effectively eliminated cancer cells as monotherapy, while minimizing leucocytosis. The optimal ADC selectively eliminated CXCR4+ cancer cells in solid tumours, but showed limited toxicity to normal CXCR4+ tissues, sparing haematopoietic stem cells and progenitors. Our work provides proof-of-concept that through empirical ADC design, it is possible to target proteins with broad normal tissue expression.

A mouse model for evaluation of efficacy and concomitant toxicity of anti-human CXCR4 therapeutics.

The development of therapeutic monoclonal antibodies through mouse immunization often originates drug candidates that are not cross-reactive to the mouse ortholog. In such cases, and particularly in oncology, drug efficacy studies are performed on human tumor xenografts or with "surrogate" anti-mouse ortholog antibodies if targeting tumor host cells. Safety assessment of drug candidate(s) is performed at a later development stage in healthy non-human primates. While the latter remains necessary before a drug advances into human subjects, it precludes evaluation of safety in disease conditions and drug de-risking during early development. Therefore, mouse models that allow concomitant evaluation of drug efficacy and safety are highly desirable. The C-X-C motif chemokine receptor 4 (CXCR4) is an attractive target for tumor-targeted and immuno-oncology therapeutics, with multiple mouse immunization-derived antibodies undergoing clinical trials. Given the pleiotropic role of CXCR4 in cancer biology, we anticipate continuous interest in this target, particularly in the testing of therapeutic combinations for immuno-oncology. Here, we describe the generation and validation of the first mouse knock-in of the whole coding region of human CXCR4. Homozygous human CXCR4 knock-in (hereafter designated as HuCXCR4KI) mice were viable and outwardly healthy, reproduced normally and nursed their young. The expression pattern of human CXCR4 in this model was similar to that of CXCR4 expression in normal human tissues. The human CXCR4 knock-in gene was expressed as a biologically active protein, thereby allowing normal animal development and adequate"homing" of leukocytes to the bone marrow. To further validate our model, we used an in vivo functional assay of leukocyte mobilization from bone marrow to peripheral blood by blocking CXCR4 signaling. Both an anti-human CXCR4 -specific blocking antibody and the small molecule CXCR4 inhibitor AMD3100 induced increased leukocyte counts in peripheral blood, whereas an anti-mouse CXCR4 -specific blocking antibody had no effect. This new mouse model is useful to evaluate efficacy and safety of anti-human CXCR4 -specific drugs as single agents or in combination therapies, particularly in the oncology, immuno-oncology, wound healing and chronic inflammation therapeutic areas.

A novel CXCR4 antagonist IgG1 antibody (PF-06747143) for the treatment of hematologic malignancies.

The chemokine receptor CXCR4 is highly expressed and associated with poor prognosis in multiple malignancies. Upon engagement by its ligand, CXCL12, CXCR4 triggers intracellular signaling pathways that control trafficking of cells to tissues where the ligand is expressed, such as the bone marrow (BM). In hematologic cancers, CXCR4-driven homing of malignant cells to the BM protective niche is a key mechanism driving disease and therapy resistance. We developed a humanized CXCR4 immunoglobulin G1 (IgG1) antibody (Ab), PF-06747143, which binds to CXCR4 and inhibits CXCL12-mediated signaling pathways, as well as cell migration. In in vivo preclinical studies, PF-06747143 monotherapy rapidly and transiently mobilized cells from the BM into the peripheral blood. In addition, PF-06747143 effectively induced tumor cell death via its Fc constant region-mediated effector function. This Fc-mediated cell killing mechanism not only enhanced antitumor efficacy, but also played a role in reducing the duration of cell mobilization, when compared with an IgG4 version of the Ab, which does not have Fc-effector function. PF-06747143 treatment showed strong antitumor effect in multiple hematologic tumor models including non-Hodgkin lymphoma (NHL), acute myeloid leukemia (AML), and multiple myeloma (MM). Importantly, PF-06747143 synergized with standard-of-care agents in a chemoresistant AML patient-derived xenograft model and in an MM model. These findings suggest that PF-06747143 is a potential best-in-class anti-CXCR4 antagonist for the treatment of hematologic malignancies, including in the resistant setting. PF-06747143 is currently in phase 1 clinical trial evaluation (registered at www.clinicaltrials.gov as #NCT02954653).